OBJECTIVE To investigate the treatment outcome for women suffering recurrent miscarriages associated with strong or moderate antiphospholipid antibody (aPL) production. METHODS Sixty-seven pregnancies in 61 women demonstrating at least one kind of aPL with a history of recurrent miscarriages were treated with: (1) aspirin (ASA) alone; (2) prednisolone (PSL) and ASA; and (3) PSL, ASA, heparin and/or immunoglobulin (IgG). For comparison purposes the aPL-positive patients were divided into two groups, strongly and moderately-positive. IgG and IgM antibodies against PE and five negatively-charged phospholipids were measured by ELISA between 1987 and 1993. Beta2-glycoprotein I (beta2GPI) dependent anticardiolipin antibodies were measured by ELISA since 1993. Lupus anticoagulant was measured by a diluted aPTT method since 1993. RESULTS Out of a total of 16 (50%) patients strongly-positive for aPL and 47 out of 51 (92.2%) moderately-positive demonstrated a successful outcome. The live birth rate moderate group was significantly higher than in the strongly-positive cases (P < 0.0005). In the cases exhibiting moderate aPL production, 28 out of 30 (93.3%) receiving PSL and ASA and 14 out of 15 (93.3%) treated with ASA alone successfully gave birth. None of the 14 given ASA alone suffered preterm delivery or IUGR. In contrast 12 (36.4%) and 6 (18.2%) of the 33 patients treated with the PSL combination therapy suffered from preterm delivery and IUGR, respectively. CONCLUSIONS The live birth rate in patients strongly positive for aPL is lower than that in patients with moderate aPL production even if treatment is performed during pregnancy. However, ASA is useful to treat cases with moderate aPL so that distinction of the two groups is warranted.